Nov 12, 2012

I survived Nodejsconf.it

On Saturday I attended and talked at the 2012 edition of Nodejsconf.it. I was pretty nervous because the talk would have been in English and... because the proposed title was "Why I hate Node.js". As stated ate the beginning of the talk I really do not hate anything. Life is too short and too complicated to hate things like web frameworks of programming languages that last only few years. But I thought also that ti would be wise, having a provocateur presentation in front of a potentially hostile crowd , to wear an helmet! The "hater" talk was inspired from a tradition I saw at the djangocon since the first edition. I do not know if it has deeper roots. I tried to explain my points with ease and humor. And it  worked. I had a very very positive feedback from both attendees and webdebs, the organizer of the conf. In the audience there were prominent personalities of the node.js community and the question time has been taught.Glenn Block, Microsoft Azure expert of Node.js.
I had also some controversial critics on joined.in about being in favor of multi-threaded programming  not showing data and not giving credits for some of the material in my slides.
So starting from the credits, I do really apologize for that. I was too easy on this and I hope this is not too little and too late.
The writer/javascript stuff was taken from the blog post if Hemingway wrote JavaScript. It appeared also in HackerMonlty after I decided to mention it in my talk ( just saying). Growing complexity examples where taken from FENN BAILEY blog post "Node js a gigant step backwards"
Something also from "is nodejs wrong" by Nicolas Cannasse. And part of the questions to make some fun of javascript typing are takent from the famous "WAT" video ( you can find it on the internet :-) ). The quote
You want to use a programming language with the word "script" in it... and "java" in the other side
Is mine.
About not showing data I tried to find some data that were not controversial. The "Hello word" example ha no meaning and the implementation could bias benchmark too much.  I displayed a chart with completely made-up data just to show how I think complexity over feature could work on a large node/javascript codebase.
I had a great time doing this talk and the majority of people seem to liked it. If I had the chance to perform it again I will try to fix it with the previous suggestions.
These are the slides



No comments: